Automated melanoma recognition

Abstract

A system for the computerized analysis of images obtained from ELM has been developed to enhance the early recognition of malignant melanoma. As an initial step, the binary mask of the skin lesion is determined by several basic segmentation algorithms together with a fusion strategy. A set of features containing shape and radiometric features as well as local and global parameters is calculated to describe the malignancy of a lesion. Significant features are then selected from this set by application of statistical feature subset selection methods. The final kNN classification delivers a sensitivity of 87% with a specificity of 92%.

A fuzzy-based histogram analysis technique for skin lesion discrimination in dermatology clinical images

“…Many color descriptors that have been applied to melanoma detection, including variation of hues [8], analytical color techniques for detecting color variegation [9], RGB color channel statistical parameters [10][11][12], spherical color coordinates and (L,a * ,b * ) color coordinate features [13], percentage of the skin lesion containing absolute shades of reddish, bluish, grayish and blackish areas and the number of those color shades present within the skin lesion [14]. Color quantization for the different color shades examined in [14] was performed using the median cut color quantization algorithm [15]. Melanoma detection has also been performed based on probability analysis for three classes of colors (benign, melanoma and other colors) found from relative color histograms [16,17].…”

Melanoma Recognition Using Representative and Discriminative Kernel Classifiers

“…Ganster et al [5] presented a system where as initial step the binary mask of the skin lesion was determined by several basic segmentation algorithms combined together with a fusion strategy [5]. The algorithms used to segment the lesion are: global thresholding, dynamic thresholding, and a 3-D color clustering concept [5].…”

“…Recently numerous research on this topic have been proposed (for a more comprehensive discussion of the most significant literature we refer the reader to section 2); a key factor for the development and evaluation of these systems is the availability of a statistically significant database. One of the largest databases of melanoma images available to the research community was contributed by H. Ganster et al [5]. That paper presented a database of 5363 images, accompanied by: (a) a segmentation algorithm for isolating the potential melanoma from the surrounding skin, determined by several basic segmentation algorithms combined together with a fusion strategy [5]; (b) a set of features containing shape and radiometric features as well as local and global parameters, calculated to describe the malignancy of a lesion, from which significant features are selected by application of statistical feature subset selection methods [5]; (c) a nearest neighbor classification algorithm [5].…”

“…One of the largest databases of melanoma images available to the research community was contributed by H. Ganster et al [5]. That paper presented a database of 5363 images, accompanied by: (a) a segmentation algorithm for isolating the potential melanoma from the surrounding skin, determined by several basic segmentation algorithms combined together with a fusion strategy [5]; (b) a set of features containing shape and radiometric features as well as local and global parameters, calculated to describe the malignancy of a lesion, from which significant features are selected by application of statistical feature subset selection methods [5]; (c) a nearest neighbor classification algorithm [5]. In that work the authors concentrated particularly on the segmentation technique and the features selection process, obtaining results that, to the best of our knowledge, represent the state-of-the-art on this topic.…”

“…We also evaluated the influence of the segmentation method by running two series of experiments: the first using the segmentation masks obtained by Ganster, the second using an hand-made rectangular mask which roughly contains the whole lesion while minimizing the amount of surrounding skin in the image. In order to have a fair comparison, we replicated the experimental setup used in [5] for a benchmark evaluation. Our results show that SVM obtains remarkably better performances than SG-MRF and Ganster's method with both feature types and regardless of the segmentation method.…”